อนุทิน 174469

Machine Intelligible (MI)

Currently new knowledge is developed and described in text and graphic narratives that require conversion and annotation to make the (new) knowledge machine intelligible (that is learn-able by digital devices). This delays inclusion of knowledge in AI framework and system. There are potential advantages for AI systems to be able to include new knowledge quickly and perhaps extend the knowledge or applications of the knowiedge by (machine) networking with other knowledge already learned (by machine).

Works in semantic nets and semantic languages and representations (eg. XML/RDF) for Web/Internet applications are showing potential for “webs of AI systems” (which may out-perform each individual AI systems).

Works to be done are a challenge for a few years ahead, but the sooner we start, the earlier we can benefit from the [super]-AI-[web] systems. To answer: How do make knowledge ‘machine intelligible’? How do we make AI system connectable or networkable? Can we verify and ‘blockchain’ knowledge topics/subjects so that the knowledge is ‘reliable’? …

Can Thailand get on this new ‘road to frontier’? Of course, we can and it may lead us to new economy and new knowledge society.